This paper introduces a method for creating naturally varied plants from a given basic plant model. Previous techniques create variation in plants by introducing local randomness to the plant model's description. However, randomness is restricted by the model's parameterization and lack of correlation between local features making varying global properties (e.g. branch and stem curvature) difficult. We present a biologically−based method which mimics the underpinnings of variation in real plants. This method uses a feedback control system to simulate the biological growth mechanism by which a plant naturally responds to environmental factors. We show that our technique creates more realistically varied models by modelling growth responses to stimuli, and provides a method for quickly creating numerous similar models, none of which are exactly alike.

We introduce a class of biologically-motivated algorithms for generating leaf venation patterns. These algorithms simulate the interplay between three processes: (1) development of veins towards hormone (auxin) sources embedded in the leaf blade; (2) modification of the hormone source distribution by the proximity of veins; and (3) modification of both the vein pattern and source distribution by leaf growth. These processes are formulated in terms of iterative geometric operations on sets of points that represent vein nodes and auxin sources. In addition, a vein connection graph is maintained to determine vein widths. The effective implementation of the algorithms relies on the use of space subdivision (Voronoi diagrams) and time coherence between iteration steps. Depending on the specification details and parameters used, the algorithms can simulate many types of venation patterns, both open (tree-like) and closed (with loops). Applications of the presented algorithms include texture and detailed structure generation for image synthesis purposes, and modeling of morphogenetic processes in support of biological research.

Abstract Because of their complexity, plant models used in computer graphics are commonly created with proceduralmethods. A difficult problem is the user control of these models: a small number of parameters is insufficient tospecify plant characteristics in detail, while large numbers of parameters are tedious to manipulate and difficultto comprehend. To address this problem, we propose a method for managing parameters involved in plant modelmanipulation. Specifically, we introduce decomposition graphs as multiscale representations of plant structuresand present interactive tools for designing trees that operate on decomposition graphs. The supported operationsinclude browsing of the parameter space, editing of generalized parameters (scalars, functions, and branchingsystem silhouettes), and the definition of dependencies between parameters. We illustrate our method by creatingmodels of bonsai trees. Categories and Subject Descriptors (according to ACM CCS): I.3.6 [Computer Graphics]: Methodology and Techniques